How to isolate RAM or motherboard in a six-month PC with random BSODs?

[AndrewJ]

I built a new gaming PC about six months ago, and it's started giving me random BSOD errors with different stop codes like IRQL_NOT_LESS_OR_EQUAL and MEMORY_MANAGEMENT, which makes pinpointing the faulty component incredibly frustrating since the crashes are intermittent and don't happen under a consistent workload. I've run the Windows Memory Diagnostic tool, updated all my drivers from the manufacturer websites, and checked my SSD health, but the problem persists without a clear pattern, and I'm worried it could be a deeper hardware issue with the RAM or motherboard. For others who have successfully resolved persistent BSOD troubleshooting, what systematic approach did you take to isolate the cause? Did you find specific logging tools more useful than the built-in Windows ones, and is there a reliable way to stress-test individual components to force the error and finally identify the culprit?

[JackP]

You're not alone. Start with a simple, repeatable test to isolate RAM first. Boot MemTest86 on a USB and run it overnight (or at least 6–8 passes). Then test each RAM module individually in its primary slot to see if the error follows a module or a slot. If RAM passes, move to CPU stress testing with Prime95 small FFTs (watch temps) or OCCT; monitor voltages with HWInfo. If that stays clean, test GPU with FurMark and check for driver conflicts. Keep a log of findings.

[Kyle30]

Collect crash dumps and analyze them: enable automatic memory dumps; use Event Viewer to find BugCheck codes; use BlueScreenView to map memory.dmp to drivers; use WinDbg to load dumps and inspect stack traces. Note which driver or module appears repeatedly. Save dumps from several incidents to spot patterns. If a driver looks implicated, try a clean boot or a driver cleanup (DDU for graphics drivers) before retesting.

[Dennis_P]

BIOS and motherboard checks can save you time. Reset BIOS to defaults, disable XMP, and reseat all RAM. Test with one stick at a time in the primary slot, then add more gradually. If available, update the motherboard BIOS to the latest version and ensure VRMs aren’t overheating. Sometimes stability improves after a microcode/BIOS update or avoiding aggressive memory timings.

[AnthonyUW]

On the hardware side, run targeted stress tests for a few hours: MemTest86 for RAM, Prime95/OCCT for CPU (watch temps and voltages), FurMark or 3DMark for GPU, and use a PSU tester if you have one or monitor rails under load with HWInfo. Don’t run everything at full tilt forever, but a coordinated test plan helps you see what triggers a crash.

[Ronald61]

Keep notes and be methodical: document every test, what changed, and the exact stop code or crash pattern. If you can, post your hardware spec (CPU, GPU, RAM, motherboard, PSU) and your OS version here so others can point to known issues or suggested tests. If you want, I can suggest a concrete week-by-week test plan tailored to your setup and budget.